In an existing standard, although a data rate of a physical layer has significantly increased, an advantage of an increase in the data rate cannot be presented at a MAC layer because of an inherent system overhead of the Media Access Control (Media Access Control, MAC for short) layer. To reduce an impact of an additional overhead, a frame aggregation mechanism is proposed in the existing standard. Essence of the frame aggregation mechanism is that a plurality of data frames shares one MAC frame header, so as to simplify a frame structure and eliminate inter-frame spaces between data frames and contention time, thereby increasing an effective throughput of the MAC layer.
Currently, two aggregation mechanisms are adopted in the standard, that is, Aggregate Medium Access Control (MAC) Service Data Unit (Aggregation-MAC Service Data Unit, A-MSDU for short) and Aggregate Medium Access Control (MAC) Protocol Data Unit (Aggregation-MAC Protocol Data Unit, A-MPDU for short).
A packet aggregation mechanism is adopted in the existing standard, so as to increase a proportion of effective load. However, when a common node aggregates data, the common node needs to wait for a certain time for the aggregated data to reach a certain scale for transmission, which largely increases a delay of the data, and obviously cannot ensure quality of service of a delay-sensitive service. Meanwhile, in a process in which a node aggregates packet data of different applications and sends aggregated packet data to an access point (Access Point, AP for short), it is difficult to ensure sensitive requirements of different applications for a delay. When a common node performs aggregation, energy consumption of the common node is severe. In an existing application scenario, for example, a smart grid, an energy problem of a node is more obvious due to a large size of the network. In a process in which a data node sends aggregated data to a destination node, a problem of conflict and collision caused by a problem of a hidden terminal is also severe.